The Murder of Children by Fathers in the Context of Child Abuseଝ

Child Abuse & Neglect 31 (2007) 731–746

The murder of children by fathers in the context of child abuseଝ

Kate Cavanagh a,∗, R. Emerson Dobash b, Russell P. Dobash b

a Department of Applied Social Science, University of Stirling, FK9 4LA Scotland, UK b Criminology, School of Law, University of Manchester, England, UK

Received 15 August 2005; received in revised form 30 October 2006; accepted 21 December 2006 Available online 12 July 2007

Abstract

Objective: This study examined the backgrounds of fathers who fatally abuse their children and the contexts within which these homicides occur. The type of relationship between victim, perpetrator, and the victim’s mother was a particular interest. Methods: Data were gathered from 26 cases of fatal child abuse perpetrated by fathers derived from the wider Murder in Britain study.** Quantitative and qualitative data were collected from extensive prison case ﬁles of men serving life sentences for child murder. Results: This was a group of undereducated, underemployed men with signiﬁcant criminal histories. All except one victim had been subjected to previous violence by the offender, almost three-quarter of whom had also perpetrated violence against their intimate partners (the child’s birth mother). Many men had unreasonable expectations and low tolerance levels of normal childhood behaviors, and many appeared jealous and resentful of these young children. All 26 victims were under 4 years of age. Sixty-two percent of the offenders were stepfathers and in only four cases was the perpetrator a birth father married to the birth mother. Stepfathers had more disrupted and disadvantaged backgrounds and experiences than birth fathers. Conclusions: Findings suggest that fathers who perpetrate fatal child abuse have a propensity to use violence against children in their care and intimate partners, raising questions about the gender dynamics and generational

ଝ Data analysis and first draft of this paper were completed when the first author was Visiting Scholar at the School of Social Work and Policy Studies at the University of Sydney, Australia. Thanks are extended to: the University of Stirling for granting sabbatical leave; colleagues in the University of Sydney for providing such a stimulating environment; and the Carnegie Foundation, Scotland for giving financial support towards this endeavour. We are also appreciative of the valuable comments received from the editor and reviewers. ∗∗ The Murder in Britain study was supported by an ESRC grant. ∗ Corresponding author.

0145-2134/$ – see front matter © 2007 Published by Elsevier Ltd. doi:10.1016/j.chiabu.2006.12.016 732 K. Cavanagh et al. / Child Abuse & Neglect 31 (2007) 731–746 boundaries operating in these families. The nature and type of intimate relationship (whether married or cohabiting) and fathering relationship (whether birth or de facto) were important differentiating factors in these homicides as well as characteristics of the offender. Practice implications: Professionals working in child protection strive to provide effective services to children and families, ever vigilant to the possibility of the death of a child as a consequence of an assault. By and large, fathers (either biological or de facro) as the perpetrators of such assaults have received minimal attention in both policy and practice. Findings from this study suggest that practitioners need to be cognisant of men’s attitudes towards and excpeectations of fathering (particularly stepfathering) which may present increased levels of risk to both children and intimate partners. © 2007 Published by Elsevier Ltd.

Keywords: Child murder; Child homicide; Fatal child; Abuse; Fathers; Stepfathers

Introduction

It is widely recognized that homicide is a male activity (Brookman & Macguire, 2003; Daly & Wilson, 1988; Mouzos, 2000; Soothill, 1999). With regard to fatal child abuse, several studies have reported a preponderance of fathers as perpetrators (Copeland, 1985; Goetting, 1988; Marks & Kumar, 1996; Somander & Rammer, 1991). Other studies have noted the prevalence of de facto fathers as perpetrators (Kaplun & Reich, 1976; Scott, 1973; Vanamo, Kauppi, Karkola, Merikanto, & Rasanen, 2001). Relatedly, a number of commentators have noted the lack of empirical data on these fathers (Adinkrah, 2003; Alder & Polk, 1996; Lucas et al., 2002) and the dominant focus on offending mothers (Corby, 2000; Victoria Child Death Review Committee, 2004). Addressing this imbalance, this paper presents ﬁndings from a study of 26 cases of fatal child abuse perpetrated by fathers (biological and de facto) convicted of the murder of their birth or stepchild. Quantitative data demonstrating patterns evident in these cases are presented, complemented with qualitative data illustrating contextual themes. The ﬁnd- ings reported here are derived from the Murder in Britain study whose aim was to provide detailed evidence about the nature, context, situations, and lethal intentions associated with different types of murder.

Background

Existing research indicates that children are most likely to be killed within the family and usually by a parent or stepparent (Alder & Polk, 2001; NSPCC, 2001; Wilczynski, 1997). Several studies have identified fatal child abuse as a distinct type of familial child homicide and the one that occurs most frequently (Blaser, 1985; Christoffel & Lui, 1983; Crittenden & Craig, 1990; Herman-Giddens, 1999; Strang, 1993). In the rich body of literature on fatal child abuse, generally accepted as the killing of a young child by a caretaker usually in the context of ongoing abuse and neglect (Creighton, 1995; De Silva & Oates, 1993; Falkov, 1996; Greenland, 1987; Hicks & Gaughan, 1995; Kotch, Chalmers, Fanslow, Marshall, & Langley, 1993; Reder & Duncan, 1999), a number of key themes have been identified including: the relationship between the perpetrator and the victim and between the perpetrator and the victim’s mother, the prevalence of previous violence to the victim and/or the victim’s mother, a range of factors associated with K. Cavanagh et al. / Child Abuse & Neglect 31 (2007) 731–746 733 the homicide itself, and the individual child and adult experiences of perpetrators. Each of these is briefly reviewed.

Types of relationships

Certain risk factors have been identified in familial child homicides, most notably the nature of the relationship between the child and the perpetrator (whether birth or de facto parent) and between the parents of the child (whether married or cohabiting). Drawing on several national homicide data sets, Daly and Wilson (1988, 1994) found that stepchildren are twice as likely to be killed in families than birth children. Arguing that de facto fathers are reluctant to invest in other men’s children, they posit that stepchildren are at greatest risk of homicide by stepfathers. Daly and Wilson (1988, 1994) also found that birth children of married relationships are less likely to be killed than in any other types of relationship. In terms of wider demographic patterns, national British data indicate that while marriage rates remain relatively stable, remarriage, divorce, and cohabitation rates have increased resulting in many more stepfamilies. Children are much more likely to stay with their biological mothers following the dissolution of an intimate relationship, and while it is difficult to gauge how many fathers there are vis-a-vis´ stepfathers, in 2003/2004, 83% of stepfamilies in Great Britain consisted of a stepfather and biological mother (Office for National Statistics, 2005).

Previous violence by perpetrator to victim and victim’s mother

Much evidence suggests that at the time of their deaths, many children killed in families have been seriously abused by the perpetrator prior to the killing (Brewster et al., 1998; Browne & Lynch, 1995; Greenland, 1987; Strang, 1995). Relatedly, in the last decade or so a developing body of research has established that some men are abusive to both their children and their partners (Edelson, 1999; Kantor & Little, 2003; Mullender et al., 2002; Rumm, Cummings, Krauss, Bell, & Rivara, 2000). Findings from a number of studies have shown that some fathers who kill their children have a history of abuse to their intimate partner (Adinkrah, 2003; Alder & Polk, 1996; Sanders, Colton, & Roberts, 1999) though this has been minimally discussed.

The homicide event

Many studies have shown that younger children, particularly those under 1 year, are more likely to be killed within the family (Blaser, 1985; Creighton, 1995; Kotch et al., 1993; Resnick, 1969). Boys and girls are equally likely to be the victims of fatal child abuse (De Silva & Oates, 1993; Donnelly, Cumines, & Wilczynski, 2001; Lucas et al., 2002), which usually take place in the home (Browne & Lynch, 1995; Marleau, Poulin, Webacnk, Roy, & Laporte, 1999), and often when the victim is in the sole and temporary care of a parent, typically the (step) father (Brewster et al., 1998; Strang, 1995). Youngerchildren are more likely to be killed by the use of direct force, and the use of blunt instruments and shaking, particularly of infants, is widely reported (Lyman et al., 2003; Wheeler, 2003). Paradoxically while the evidence indicates that much sexual abuse occurs within families (Bagley & Thurston, 1996; Leventhal, 1998), sexual assault in familial child killings more broadly and in fatal child abuse speciﬁcally is rarely reported (Donnelly et al., 2001; Jason & Andereck, 1983; Pollanen, Smith, Chiasson, Cairns, & Young,2002; Wilczynski, 1995). 734 K. Cavanagh et al. / Child Abuse & Neglect 31 (2007) 731–746

Contextual aspects of child abuse homicides

Where motive and intention is examined, studies of fatal child assault suggest that male perpetrators: often overestimate their child care abilities and become frustrated at their inability to elicit the desired response from a child (Greenland, 1987); use harsh and punitive disciplining of a child for behavior regarded as annoying or disobedient (Adinkrah, 2003; Somander & Rammer, 1991); demonstrate an intolerance of aspects of the child’s behavior, for example, crying (Baker, Craig, & Lonergan, 2003; Kasim, Cheah, & Shaﬁe, 1995; Shepherd & Sampson, 2000). Several studies suggest that these homicides lack “intentionality” and that men’s motivation is to discipline rather than kill the child (Crittenden & Craig, 1990; Resnick, 1969; Stroud & Pritchard, 2001). However, the need to focus on the presence of intention to harm rather than on the apparent absence of intention to kill the victim is highlighted by other studies which suggest that child killings perpetrated by fathers may be motivated by anger against an intimate partner displaced to the child and may involve jealousy of or resentment toward the child for occupying the mother’s attention (Alder & Polk, 2001; Wilczynski, 1995).

Child and adult background of offenders

Research in child development and criminology has consistently shown that problems experienced within the family and the quality and stability of caring are linked to child and adult offending behavior (Loeber & Stouthamer-Loeber, 1998; Nagin & Farrington, 1992; Shaw & Winslow, 1997). The intergenerational transmission of abuse has been a continuous theme in the child abuse literature (Cicchetti & Aber, 1980; Hunter & Kilstrom, 1979; Kaufman & Zigler, 1987; Kempe, Silverman, Steele, Droegemueller, & Silver, 1962), and while findings are equivocal, the view that experience of abuse as a child is a strong indicator of future abuse continues to be dominant. Evidence of the intergenerational transmission of abuse in relation to child homicide is, however, limited (Lucas et al., 2002). Poverty, unemployment, and minimal education are perennial features in the lives of all types of child killers, including those who fatally abuse children (Campion, Cravens, & Covan, 1988; Schloesser, Pierpont, & Poertner, 1992; Strang, 1995). Problem behaviors such as alcohol and/or drug abuse (Somander & Rammer, 1991)or a history of offending, particularly involving violence are reported (Alder & Polk, 1996; Donnelly et al., 2001; Wallace, 1986). A history of mental health problems has also been highlighted (Falkov, 1996; Trocme & Lindsey, 1996). In summary, the literature identifies a number of key themes associated with familial child homicide. Importantly, the need for more knowledge about fathers who fatally abuse their children and consideration of the status of the fathering relationship have been identified. Using both quantitative and qualitative data, the latter seldom reported in the literature (Reder, Duncan, & Gray, 1993), this study examines the backgrounds and experiences of fathers who fatally abused their children and the contexts within which such crimes were perpetrated.

Methods

The Murder in Britain study was a 3-year investigation of all types of murder. The purpose was to conduct an intensive study of murder in order to provide detailed information regarding a range of theoretically derived constellations of factors linked to lethal violence. These included the nature K. Cavanagh et al. / Child Abuse & Neglect 31 (2007) 731–746 735 of the relationship between victim and offender, childhood and adult backgrounds, criminal careers, and the nature and contexts of lethal acts. All the necessary ethical approvals were secured from the Home Office and Scottish Executive prior to the study commencing. A human subject review was not required. Data were gathered from three distinct sources: national Homicide Indices for England/Wales and Scotland, which contain a limited amount of data; case files of a sample of 866 men and women convicted of murder; and 200 in-depth interviews with men and women serving life sentences in prison for murder (Lewis, Dobash, Dobash, & Cavanagh, 2003). Here both quantitative and qualitative data extracted from the case files are presented.

Case ﬁle data set

The case file data set was comprised of cases resulting in a conviction for murder which leads to an automatic life sentence in prison accompanied by a recommendation outlining the minimum period to be served. Because of the gravity of the offence of murder and the “indeterminate” nature of the life sentence, the prison services invest considerable effort in dealing with those convicted of murder who must be judged “safe” before being released into the community. Case files are substantive documents containing official reports from the different professionals required to assess and provide evidence concerning the offender and the homicide event, for example, police, pathologists, trial judges, psychiatrists, probation officers, social workers, and prison staff. Each file contains extensive information on the childhood and adulthood of offenders, circumstances surrounding the homicide and the event itself, the legal process, and activities and adjustments in prison. The files also include accounts from perpetrators. Analysis of this comprehensive data concerning different types of murder allowed us to explore contexts, situations, and intentions (Cavanagh, Dobash, & Dobash, 2005; Dobash, Dobash, Cavanagh, & Lewis, 2004; Dobash, Dobash, Cavanagh, & Medina-Ariza, 2007). To maximize this source of extensive information, case files in both Scotland and England were read to establish the nature and consistency of their content; an initial data collection instrument (containing over 400 quantitative variables) and coding frame were developed that allowed data to be coded directly onto laptop computers using SPSS. The instrument was piloted in both countries and modified. Throughout the data collection period, coding queries were discussed by the research team. These procedures maximized reliability and validity and produced a robust data set containing little missing data. Qualitative data were also gathered from the case files and analyzed using QSR.NUD*IST. There were 866 cases in the case file data set, 786 men and 80 women. Files for all male offenders currently serving life sentences for murder are situated in the respective prison headquarters in London (for England/Wales) and Edinburgh (for Scotland). During the years 1999 and 2000, a systematic sample of 786 cases was selected from each universal archive along with all cases from several prisons identified as strategic sites where interviews were to be conducted. Overall this sample represents about 20% of all men currently in prison for murder (of any type) in England and Wales (about 3000 men) and 35% of those similarly sentenced in Scotland (about 500 men). Of the 786 cases of murder perpetrated by men, 90 (11%) involved the murder of a child. There were 41 cases of nonfamilial child murder and 49 cases of familial child murder. Included in the latter were 26 cases of fatal child abuse. The vast majority of these cases occurred in the 1990s. The literature indicates that fatal child abuse is perpetrated predominantly against children under 5 years of age. In England, Wales, and Scotland during that decade there were approximately 34 deaths of children under the age of 5 recorded annually as homicides (Cotton, 2003; 736 K. Cavanagh et al. / Child Abuse & Neglect 31 (2007) 731–746

Scottish Executive, 2001). Such cases included those where either parent was the perpetrator and not all resulted in a conviction for murder. Fatal child abuse has been previously defined as “...the killing of a child by one or a series of assaults by a parent or a person with the status of a parent to eliminate a disturbing behavior of a child without the intention to kill” (Somander & Rammer, 1991, p. 47). Adapting this definition, fatal child abuse was defined as those cases involving the murder of a child by a birth or de facto father to eliminate a disturbing behavior where the child was the primary target of the perpetrator’s actions. The approach taken here differs in that it incorporates the notion that these acts are purposeful and that intentionality is an important aspect of understanding why men fatally abuse their children. These cases constituted over a quarter (29%) of all the child killings perpetrated by men in the Murder in Britain study and over half (53%) of the familial child killings. The remaining 47% of the latter were excluded as they did not fit the above definition, and included cases where, for example, the male perpetrator was related to, but was not the father of the victim. Similar to national data sets (e.g., Office for National Statistics, 2005), a child was defined as a dependent person, under 16 years of age and living in a household. British legislation assumes that adult status is attained at 16 years when young people can legally leave school, commence paid employment, and marry (in Scotland). “Cohabitation” was defined as a non state-sanctioned intimate relationship between a man and a woman who live together in the same household with any children, and “stepfather” was defined as a man, either married to or cohabiting with a woman, who is acting as a parent to the biological children of his intimate partner. Aggregate data outlining similarities and differences between biological fathers and stepfathers across a number of variables relating to the childhoods and adulthoods of the two groups are presented. Fisher exact test was used for comparison. Qualitative data taken from the case files are used to illustrate a number of key contextual themes to emerge from an analysis of the case file accounts, including the nature of the murder incident, previous violence to the victim and intimate partner, and motivations and intentions. These data are not viewed as “fact” but neither are the information, ideas and interpretations of events viewed as “fiction” with no objective reality. Such documents deal with a social and material reality that has meanings, impacts and consequences and the data contained within these narratives are invaluable in developing our understanding of some of the key contextual elements of these crimes.

Results

The 26 child victims were between 3 weeks and 4 years old (mean of 16 months). There were no second child victims or adult victims. Half of the victims were girls. All except one of the murder incidents occurred in the home of the victim and perpetrator. There were coaccuseds in three cases and eye-witnesses in two, all of whom were the birth mothers of the victim and intimate partner of the perpetrator. At the time of the murder, the perpetrators were aged 21–32 (mean of 25 years).

Relationship between victim and offender

Of the 26 victims, 62% were stepchildren, and 38% were birth children. Twenty-one of the intimate relationships were cohabitations (81%), four (15%) were marriages, and one (4%) was a dating relationship. In three of the four marriages, the perpetrator was the birth father. Of the 21 cohabitations, 71% K. Cavanagh et al. / Child Abuse & Neglect 31 (2007) 731–746 737

Table 1 Biographical details of childhood of birth and stepfather perpetrators Percent of birth Percent of stepfathers Total (N = 26) fathers (N = 10) (N = 16) Problems in family and stability of care At least three carers 20 36 29 Parental relationship breakdown 20 64* 46 Parental alcohol abuse 10 31 22 Parental drug abuse 0 0 0 Parental mental health problems 0 21 13 In care as child 10 29 21 Experience of violence and abuse Father violent to mother 13 33 25 Physical violence by carers 10 31 22 Sexual abuse 0 8 4 Problem and offending behavior Problems at school 20 73* 48 Alcohol abuse 40 75 62 Drug abuse 0 67** 40 Mental health problems 10 25 19 Offending as a child 20 50 38 At least six convictions 20 21 21 Violence to others as child 11 14 13 * p < 0.05 ** p < 0.01. were stepfathers and 29% were birth fathers. In all 26 cases, the victim was living with the birth mother; stepmothers did not feature in any of these cases. Of the 10 birth fathers, 60% killed children under 6 months of age, whereas 6% of the 16 stepfathers did so. Of the 16 stepchildren killed, 94% were between the ages of 1 and 4 years. Table 1 shows the prevalence of fragmented and disruptive childhood experiences in the lives of both birth and stepfathers. When the two groups are compared, a more nuanced picture emerges. There are three statistically significant differences between the two groups of perpetrators: over three times as many stepfathers had experienced the breakdown of their parents’ relationship and had problems at school; and whereas none of the birth fathers had abused drugs, two-thirds of the stepfathers had. Overall, findings in Table 1 show that across all of the variables presented except one, the proportion of stepfathers is higher, indicating that these men appear to have had more disrupted, disadvantaged and problematic childhoods than birth fathers. As shown in Table 2 which presents some biographical features of adulthood, this was a group of undereducated and underemployed men, with limited financial resources and significant criminal histories. Marriage featured in the lives of less than one-quarter of these men, the majority of whom had a history of using violence to others especially intimate partners. Ninety-six percent of offenders had been violent to the child victim on at least one occasion prior to the murder incident. Three significant differences between the two groups of offenders emerged: over five times as many stepfathers had problems with alcohol abuse; all the stepfathers had at least one conviction whereas just under two-thirds of the birth 738 K. Cavanagh et al. / Child Abuse & Neglect 31 (2007) 731–746

Table 2 Biographical details of adulthood of birth and stepfather perpetrators Percent of birth Percent of stepfathers Total (N = 26) fathers (N = 10) (N = 16) Education, employment, and problem and offending behaviors Education at least GCE/SCE (16 years)a 20 27 24 Regularly employed 44 20 29 Alcohol abuse 10 53* 36 Drug abuse 20 47 36 Mental health problems 10 13 12 ≥1 prior convictionb 60 100** 85 ≥6 prior convictionsb 30 80* 60 ≥1 prior conviction: serious assaultc 20 44 35 ≥1 prior conviction: sexual assaultd 01912 Prior prison sentence 20 50 39 Intimate relationships and violence Ever married 30 19 23 Ever cohabited 80 93 89 Birth children 100 81 89 Previous violence by perpetrator to victim 90 100 96 Violence by perpetrator to partner (victim’s mother) 67 73 71 Violence by perpetrator to previous intimate partner 44 73 63 a GSE is the “General Certificate in Education” applicable in England/Wales and SCE is the equivalent “Standard Certificate in Education” pertaining in Scotland. b Not including childhood. c Serious assault includes convictions for any of the following indictments: attempted murder; grievous bodily harm; actually bodily harm; aggravated assault; serious assault; assault to severe injury. d Sexual assault refers to convictions for: rape; attempted rape; indecent assault. * p < 0.05. ** p < 0.01. fathers had, and more than twice as many stepfathers had at least six convictions. Importantly, Table 2 reveals that only stepfathers (19%) had convictions for sexual assault. Overall, findings in Table 2 reveal that across all of the variables presented except three (in regular employment, ever married, had birth children), the proportion of stepfathers is higher indicating that in adulthood, as in childhood, stepfathers tended to have more problematic and “nontraditional” lives than birth fathers.

The murder incident: contexts, situations, and intentions

In this section, qualitative data are used to illustrate themes identiﬁed in the reports of professionals with formal involvement in the case and also reﬂected in the literature.

Nature and context of violence. In the vast majority of cases, the victims were punched, beaten, kicked, hit, and/or swung against something. Shaking was a notable feature of many assaults, some post mortem reports noting that victims had, “...literally been shaken to death” (1087cf). In one case, a 3-month-old stepdaughter was drowned in the bath; there were no reports of poisoning, suffocation, or the use of K. Cavanagh et al. / Child Abuse & Neglect 31 (2007) 731–746 739 weapons. Sexual assault, perpetrated by a stepfather, was reported in one case. The ferocity of many of the attacks was noteworthy. 1124cf: (Trial Judge Report) “The defendant murdered his 15 month old son by a single blow of such force that it caused laceration of the child’s liver and consequent death.” 271cf: (Perpetrator) “He had this tantrum and my hand came down and I swung out and hit him hard, very hard. It must’ve hurt and he flew backwards and landed over near the fireplace.” An important contextual element of many of these fatal assaults was that the victim was in the sole care of the perpetrator, mothers having left fathers with temporary responsibility for the care of the child. The data indicate some men appeared unable or unwilling to discharge their limited parenting responsibilities without resorting to violence. When faced with a demanding and/or distressed child, offenders had choices. They potentially had a range of problem-solving behavior to be used when faced with difficult situations. Findings indicated that violence appeared to be a preferred choice for many offenders and their propensity to use violence to resolve problems was highlighted by trial judges. 095cf: (Trial Judge Report) “On his own admission, (perpetrator) gained a lot of satisfaction with his abuse of power and this abuse centred on his violence. This is, in my opinion, one of the major risk factors in this case and will have to be addressed during the custodial process.” The use of violence as a means of exercising control and abusing power is cited above as a “major risk factor.” To what extent, then, was the use of previous violence against either the victim or the victim’s mother (partner) a feature of familial life?

Previous violence against the victim. Quantitative findings presented earlier revealed that 25 of these 26 children (96%) had been previously physically assaulted by the perpetrator. In 61% of cases social services had been involved. In the single case where no previous violence was reported, the first assault was a fatal one. 951cf: (Trial Judge Report) “When the child’s body was examined, some 52 bruises were found, all inflicted in the week or so before death.” 1028cf: (Trial Judge Report) “This child was savagely beaten in an onslaught of blows some of which killed her. There was strong evidence that he had frequently used violence on the child in the last year of her life.” The above quotations illustrate, once again, not only the ferocity of the violence perpetrated against the victims, but also its frequency.

Previous violence against intimate partner. Findings presented in Table 2 showed that violence against the victim’s mother was occurring in 71% of intimate relationships. The following quotes illustrate the pattern: 1134cf: (Probation Report) “...he continued to treat his wife with increasing violence. Eventually her fear of him undermined her ability to report his violence. The attacks on her resulted in her 740 K. Cavanagh et al. / Child Abuse & Neglect 31 (2007) 731–746

nose being broken, probably on more than one occasion, her legs being cut with a Stanley knife and cigarette burns.” 951cf: (Trial Judge Report) “The relationship of the father and the mother was very stormy and the background was that this defendant had treated her with violence on many, many occasions.” The frequency and severity of the violence perpetrated by men against intimate partners is illustrated above. The case ﬁles revealed that many women lived in fear of their partners and that violence and abuse against a partner and a child often coexisted.

Jealousy and resentment. Many of the perpetrators appeared to be resentful and were jealous of these children. This derived in part from a tendency to see children as rivals for the affection and attention of their partners. The following quotes illustrate this. 854cf: (Probation Report) “He (perpetrator) explained that he was jealous that his cohabitee (mother of victim), was going out without him and that he was left at home to babysit. Letters show his infatuation with x (victim’s mother) and his frustration at being neglected.” 095cf: (Trial Judge Post-Trial Report) “There was a suggestion that he [perpetrator] was resentful of the amount of time which her mother spent on her [victim].”

Motivations. Alongside the recurring themes of jealousy and resentment, the behavior of the child was frequently identified as the “trigger” which preceded the fatal assault. 976cf: (Perpetrator) “He kept crying and I kept trying to comfort him. I’m walking around and getting more and more wound up and wondering why doesn’t he just shut up and I shook him like a rag doll and threw him on the floor.” 1134cf: (Probation Report) “He (defendant) told me he found the emotional demands of the baby difficult and said the baby’s crying, ‘...got on my nerves’.” 1028cf: (Trial Judge Report) “The perpetrator attempted to ‘discipline’ the child because she wet the bed.” As illustrated above, crying was often the apparent antecedent of the fatal assault. Some men found themselves unable to pacify their distressed charges, and a crying baby who might have been soothed by nonviolent techniques, was ultimately silenced by force. A variation on this theme was the older child (usually a toddler) who misbehaved in ways the perpetrators found displeasing. 1087cf: (Perpetrator) “I was undressing (victim) for the bath and children being children he started messing about and I ended up slapping and hitting him on the head.” A related theme was the offender’s knowledge and expectations of children. 104cf: (Trial Judge Report) “He (perpetrator) fully admits that he did not know how to parent a child. Having had no experience of young children, he feels that he treated his son as being older than he was.” 1124cf: (Trial Judge Report) “... the evidence tended to show that his expectations of the child responding to commands etc were unrealistic.” K. Cavanagh et al. / Child Abuse & Neglect 31 (2007) 731–746 741

Some offenders were particularly challenged by aspects of a child’s behavior, and their interactions with the child reflected minimal parenting skills and minimal knowledge and understanding of child development. What this raises is not only the level of commitment some men have to parenting but also the ways in which they conceive of parenting itself, as reflected in the following quotation. 595cf: (Perpetrator) “I didn’t grab her really tight but I didn’t really go mad, you know. I just wanted her to take some notice of me.” The above offender’s wish that the child respond to his need to be “noticed” illustrates the extent to which some fathers viewed children as adults with responsibilities toward them rather than the reverse. This also reflects a lack of empathy with the child.

Intentionality. The issue of intentionality while related to motivation is complex and merits separate examination. Consider the following quote: 1124cf: (Trial Judge Report) “...the defendant’s intent had been to cause really serious harm but not to kill ...this is a violent young man who had used violence upon the mother of the child and the child himself on numerous occasions prior to the act which caused the child’s death, this violence being both controlled and uncontrolled.” The above quotation not only highlights prior violence to the victim and his mother but the distinction made between intention to harm and intention to kill. Findings from this study revealed that the motivation underlying many fatal assault/s was to silence and/or to punish the child for her/his behavior. As illustrated below, few men apparently intended to kill the victim. 188cf: (From pre-sentence psychiatric report). “(Perpetrator) states that it was in sheer frustration that he hit the crying child a couple of times; something ‘flipped,’ and he did not realize that the force he used would be sufficient to kill the baby.” 947cf: (Perpetrator) “I only pushed on his chest when he wasn’t breathing. I never intended to kill him.” Revealed above are ways in which some offenders sought to mitigate their culpability for the fatal assault by minimising its severity and denying intentionality. Professionals also regularly commented on intentionality. 1018cf: (Trial Judge Report) “The blow which caused the death appears to have been impulsive and unpremeditated.” 854cf: (Trial Judge Report) “When further questioned, (perpetrator) explained that the assault had taken place because he had lost his temper. He did not intend to hurt the child, but knew that he probably would by his actions.” While some men may not have explicitly planned to kill a child, their abusive behaviors prior to the event indicated an intention to discipline and/or punish the child through violence and, by implication, the exercise of choice. If such choices had previously included violence perpetrated against their children, which happened in all but one of these 26 cases, then definitions and descriptions of these killings as “unintentional” would appear to ignore the obvious purposefulness of such acts and mask the meaning of these acts to perpetrators. 742 K. Cavanagh et al. / Child Abuse & Neglect 31 (2007) 731–746

Discussion

Key findings from this study suggest a number of risk factors that are both distinct and interrelated. The majority of men (62%) were not the biological father of the victim, and the great majority were cohabiting with (81%) rather than married to (15%) the birth mother. These findings highlight the importance of relationships, both between the perpetrator and the child and between the perpetrator and his intimate partner. Other findings point to the importance of individual experiences of perpetrators including fragmented and disruptive childhoods, and adulthoods characterized by minimal education, persistent unemployment, criminal convictions, and a history of violence. In terms of parenting practices, qualitative analysis indicated that many men had unreasonable expectations and low tolerance levels of normal childhood behaviors; some viewed children as adults with responsibilities towards them; and many were jealous and resentful of the child. While few men intended to kill the victim, the motivations of many were clear: they wanted to silence, punish, and/or discipline the child. Men’s violence towards the victim and the rationalizations they used to mitigate culpability reflected their belief that their right to attention, to silence, to time from an intimate partner had been infringed by the child. Underpinned by notions of entitlement, these attitudes also influenced their behavior toward their intimate partner, with two-thirds of the cases involving violence against the child’s mother. Findings provide support for Daly and Wilson’s (1988) argument that some stepfathers are reluctant to invest in other men’s children. Evidence suggested that many men had a limited relationship with the child, and the vast majority did not commit via marriage to the child’s mother. Many offenders seemed to view the child as a nuisance and/or an impediment to their own self-interests. In this respect, the concept of stepfather, a term used throughout to denote the nonbiological father, may need to be reconsidered in light of issues such as differential commitment to and participation in the social and emotional aspects of fathering. The over-representation of stepfathers and cohabitants suggests the importance of considering both of these types of relationships in preventive strategies. In this respect, the American Law Institute’s proposals to enhance the legal right of stepparents or ex live-in lovers by conferring custody and visitation rights (Wilson, 2006) needs careful consideration given the results presented here. While a range of problems characterized the childhoods and adulthood of the entire sample, they were more prevalent across the life course of stepfathers. In a recent examination of domestic violence, Brownridge and Halli (2002) suggest that different rates of domestic violence associated with married and cohabiting couples may best be understood in terms of the characteristics of the individuals in each category rather than the nature of the relationship itself. To the extent that in this study stepfathers had more problematic background than birth fathers, then these experiences might account for the higher likelihood of violence rather than the status of the relationship itself. However, while both sets of risk factors might be significant, sorting out their relative importance is complex, and a study of this size can only be suggestive. The focus on the supposed absence of intention to kill evident in much of the literature, minimizes the obvious purposefulness of such acts and the full meaning of intention in British law whereby intent may be established if it can be proved that the offender has disregarded the consequences of their actions (Ashworth & Mitchell, 2000; Macdonald, 1948). While many offenders may not intend to kill, they may intend to demonstrate their authority over the victim by using violence to achieve their own wishes, that is, that a child “do as they are told.” Consideration of this wider meaning of intentionality alongside an established propensity to use violence against the child, and the other risk factors identified in this study, may provide a more subtle, complex, and realistic basis for assessing the potential for lethality. K. Cavanagh et al. / Child Abuse & Neglect 31 (2007) 731–746 743

There are several limitations to this study. Findings are based on an analysis of a small sample of 26 British men serving a life sentence in prison following a conviction for the murder of a birth or stepchild. Cases of men convicted of manslaughter or culpable homicide were excluded. The data presented here are derived solely from information contained within official case files reporting the observations of a number of different professionals and often compiled over a considerable period of time. The results are therefore subject to the limitations associated with historical and retrospective information gathered from several sources. Future preventative policies may be informed by key findings from this study indicating an elevated risk of child abuse and lethality among men with problematic personal histories who are not the biological father of the child and who cohabit with rather than marry the mother of the child. Findings which raise questions about the nature and extent of some men’s participation in and expectations of parenting and their purposeful use of violence to control children and intimate partners might inform the development of protective policies and services. Given the links identified in the literature between fatal and non-fatal child abuse, findings presented here may inform an understanding of non-fatal events.

Conclusion

The majority of offenders in this study were stepfathers cohabiting with birth mothers; the nature of the intimate relationship between the perpetrator and the child’s birth mother (whether married or cohabiting), between the offender and the victim (whether birth or stepfather) were important differentiating factors in these homicides. Many offenders had an established propensity to use violence against children in their care and intimate partners, raising questions about the gender dynamics and generational boundaries operating in these families. Fundamental prevention strategies need, as a priority, to engage men in the role and task of fathering the children with whom they live, whether they are their own biological children or their stepchildren.

Acknowledgments

Thanks are extended to colleagues in the University of Sydney for offering such a stimulating working environment. We are also appreciative of the valuable comments received from the editor and reviewers.

References

Adinkrah, M. (2003). Men who kill their own children: Paternal ﬁlicide incidents in contemporary Fiji. Child Abuse & Neglect, 27, 557–568. Alder, C., & Polk, K. (1996). Masculinity and child homicide. British Journal of Criminology, 36(3), 396–411. Alder, C., & Polk, K. (2001). Child victims of homicide. Cambridge: Cambridge University Press. Ashworth, A., & Mitchell, B. (2000). Rethinking English homicide law. Oxford: Oxford University Press., 1–20. Bagley, C., & Thurston, W. (1996). Understanding and preventing child sexual abuse. Children: Assessment, social work and clinical issues and prevention education Aldershot: Arena Social Work Series. Baker, A., Craig, B., & Lonergan, G. (2003). Homicidal commotio cordis: The ﬁnal blow in a battered infant. Child Abuse & Neglect, 27, 125–130. 744 K. Cavanagh et al. / Child Abuse & Neglect 31 (2007) 731–746

Blaser, M. (1985). Epidemiologic characteristics of child homicides in Atlanta 1970–1980. Pediatrician, 12, 63–67. Brewster, A., Nelson, J., Hymel, K., Colby, D., Lucas, D., McCanne, R., & Milner, J. (1998). Victims, perpetrator, family and incident characteristics of 32 infant maltreatment deaths in the United States Air Force. Child Abuse & Neglect, 22, 91–101. Brookman, F., & Macguire, M. (2003). Reducing homicide: A review of the possibilities. Occasional Paper 84. London: Home Office. Browne, K., & Lynch, M. (1995). The nature and extent of homicide and fatal abuse. Child Abuse Review, 4, 309–316. Brownridge, D., & Halli, S. (2002). Understanding male partner violence against cohabiting and married women: An empirical investigation with a synthesized model. Journal of Family Violence, 17, 341–361. Campion, J., Cravens, J., & Covan, F. (1988). A study of filicidal men. American Journal of Psychiatry, 145, 1141–1144. Cavanagh, K., Dobash, R. E., & Dobash, R. P. (2005). Men who murder children inside and outside the family. British Journal of Social Work, 35, 667–689. Christoffel, K., & Lui, K. (1983). Homicide in childhood in 23 developed countries: US rates atypically high. Child Abuse & Neglect, 7, 339–345. Cicchetti, D., & Aber, L. (1980). Abused children-abusive parents: An overstated case. Harvard Educational Review, 50, 244–255. Copeland, A. (1985). Homicide in childhood. The American Journal of Forensic Medicine and Pathology, 6, 21–24. Corby, B. (2000). Child abuse: Towards a knowledge base. Buckinghamshire: Open University Press. Cotton, J. (2003). Homicide. In C. Flood Page & J. Taylor (Eds.), Crime in England and Wales 2001/2002: Supplementary volume. London: Home Office Statistical Bulletin. Creighton, S. (1995). Fatal child abuse—How preventable is it? Child Abuse Review, 4, 318–328. Crittenden, P., & Craig, S. (1990). Developmental trends in the nature of child homicide. Journal of Interpersonal Violence, 5, 202–216. Daly, M., & Wilson, M. (1988). Homicide. New York: Aldine de Gruyter. Daly, M., & Wilson, M. (1994). Some differential attributes of lethal assaults on small children by stepfathers versus genetic fathers. Ethology and Sociobiology, 15, 207–217. De Silva, S., & Oates, R. (1993). Child homicide—The extreme of child abuse. Medical Journal of Australia, 158, 300–301. Dobash, R. E., Dobash, R. P., Cavanagh, K., & Lewis, R. (2004). Not an ordinary killer—Just an ordinary guy. When men murder an intimate woman partner. Violence Against Women, 10, 577–605. Dobash, R. E., Dobash, R. P., Cavanagh, K., & Medina-Ariza, J. (2007). Lethal and non-lethal violence against women in intimate relationships. Violence Against Women, 13, 329–353. Donnelly, H., Cumines, S., & Wilczynski, A. (2001). Sentenced homicides in New South Wales. Sydney: Judicial Commission of New South Wales. Edelson, J. (1999). The overlap between child maltreatment and woman battering. Violence Against Women, 5, 134–154. Falkov, A. (1996). Study of working together ‘Part 8’ reports: Fatal child abuse and parental psychiatric disorder: An analysis of 100 area child protection committee case reviews under the 1989 Children Act. London: Department of Health. Goetting, A. (1988). When parents kill their young children: Detroit 1982–1986. Journal of Family Violence, 3, 339–346. Greenland, C. (1987). Preventing CAN deaths: An international study of deaths due to child abuse and neglect. London: Tavistock. Herman-Giddens, M. (1999). Under ascertainment of child abuse mortality in the United States. Journal of the American Medical Association, 28, 463–467. Hicks, R., & Gaughan, D. (1995). Understanding fatal child abuse. Child Abuse & Neglect, 19, 855–863. Hunter, R., & Kilstrom, N. (1979). Breaking the cycle in abusive families. American Journal of Psychiatry, 136, 1320–1322. Jason, J., & Andereck, N. (1983). Fatal child abuse in Georgia: The epidemiology of severe physical child abuse. Child Abuse & Neglect, 7, 2–7. Kantor, K., & Little, L. (2003). Defining the boundaries of child neglect. When does domestic violence equate with parental failure to protect? Journal of Interpersonal Violence, 18, 338–355. Kaplun, D., & Reich, R. (1976). The murdered child and his killers. American Journal of Psychiatry, 133, 809–813. Kasim, M., Cheah, I., & Shafie, H. (1995). Childhood deaths from physical abuse. Child Abuse & Neglect, 19, 847–854. Kaufman, J., & Zigler, E. (1987). Do abused children become abusive parents? American Journal of Orthopsychiatry, 57, 186–192. Kempe, C., Silverman, F., Steele, B., Droegemueller, W., & Silver, H. (1962). The battered child syndrome. Journal of the American Medical Association, 181, 17–24. Kotch, J., Chalmers, D., Fanslow, J., Marshall, S., & Langley, D. (1993). Morbidity and death due to child abuse in New Zealand. Child Abuse & Neglect, 17, 233–247. K. Cavanagh et al. / Child Abuse & Neglect 31 (2007) 731–746 745

Leventhal, J. (1998). Epidemiology of sexual abuse of children: Old problems, new directions. Child Abuse & Neglect, 22, 481–1481. Lewis, R., Dobash, R. P., Dobash, R. E., & Cavanagh, K. (2003). Researching homicide: Methodological issues in the exploration of lethal violence. In R. Lee & B. Stanko (Eds.), Researching violence: Essays on methodology and measurement (pp. 49–65). London: Routledge. Loeber, R., & Stouthamer-Loeber, M. (1998). Development of juvenile aggression and violence: Some misconceptions and controversies. American Psychologist, 53, 242–259. Lucas, D., Wezner, K., Milner, J., McCanne, T., Harris, I., Munroe-Posey, C., & Nelson, J. (2002). Victim, perpetrator, family and incident characteristics of infant and child homicide in the United States Air Force. Child Abuse & Neglect, 26, 167–186. Lyman, J., McGwin, G., Malone, D., Taypor, A., Brissie, R., Gregroy, D., & Rue, L. (2003). Epidemiology of child homicide in Jefferson County in Alabama. Child Abuse & Neglect, 27, 1063–1073. Macdonald, J. (1948, Reprinted 1986). A practical treatise on the criminal law of Scotland (5th ed.). Edinburgh: W. Green. Marks, M., & Kumar, R. (1996). Infanticide in Scotland. Medicine, Science and the Law, 36, 299–305. Marleau, J., Poulin, B., Webacnk, T., Roy, R., & Laporte, L. (1999). Paternal filicide: A study of 10 men. Canadian Journal of Psychiatry, 44, 57–63. Mouzos, J. (2000). Homicidal encounters: A study of homicide in Australia 1989–1999. Canberra: Australian Institute of Criminology. Mullender, A., Hague, G., Imam, U., Kelly, L., Malos, E., & Regan, L. (2002). Children’s perspectives on domestic violence. London: Sage. Nagin, D., & Farrington, D. (1992). The onset and persistence of offending. Criminology, 30, 501–524. National Society for the Prevention of Cruelty to Children. (2001). Out of sight: NSPCC report on deaths from abuse: (1973–1988). London: NSPCC. Pollanen, M., Smith, C., Chiasson, D., Cairns, J., & Young, J. (2002). Fatal child abuse-maltreatment syndrome: A retrospective study in Ontario, Canada 1990–95. Forensic Science International, 126, 101–104. Reder, P., & Duncan, S. (1999). Lost innocents: A follow up study of fatal child abuse. London: Routledge. Reder, P., Duncan, S., & Gray, M. (1993). Beyond blame: Child abuse tragedies revisited. London: Routledge. Resnick, P. (1969). Child murder by parents: A psychiatric view of filicide. American Journal of Psychiatry, 126, 325–333. Rumm, P., Cummings, P., Krauss, M., Bell, M., & Rivara, F. (2000). Identified spouse abuse as a risk factor for child abuse. Child Abuse & Neglect, 24, 1375–1381. Sanders, R., Colton, M., & Roberts, S. (1999). Child abuse fatalities and cases of extreme concern: Lessons from reviews. Child Abuse & Neglect, 23, 257–268. Schloesser, P., Pierpont, J., & Poertner, J. (1992). Active surveillance of child abuse fatalities. Child Abuse & Neglect, 16, 3–10. Scott, P. (1973). Fatal battered baby cases. Medicine, Science and Law, 13, 197–206. Scottish Executive. (2001). Statistical bulletin. Criminal Justice Series. Homicide in Scotland 2000. Edinburgh: Scottish Exec- utive. Shaw, D., & Winslow, E. (1997). Precursors and correlates of anti-social behavior from infancy to pre-school. In D. Stoff, J. Breiling, & J. Maser (Eds.), Handbook of anti-social behavior (pp. 37–56). New York: John Wiley. Shepherd, J., & Sampson, A. (2000). Don’t shake the baby: Towards a prevention strategy. British Journal of Social Work, 20, 721–735. Office for National Statistics. (2005). Social trends. No. 35. London: HMSO. Somander, L., & Rammer, H. (1991). Intra and extra familial child homicide in Sweden 1971–1980. Child Abuse & Neglect, 15, 45–55. Soothill, K. (1999). Homicide in Scotland. Edinburgh: Scottish Executive. Strang, H. (1993). Homicides in Australia, 1991–2. Canberra: Australian Institute of Criminology. Strang, H. (1995). Child abuse homicides in Australia: Incidence, circumstances, prevention and control. In D. Chappell & S. Egger (Eds.), Australian violence, contemporary perspectives (pp. 71–86). Canberra: Australian Institute of Criminology. Stroud, J., & Pritchard, C. (2001). Child homicide, psychiatric disorder and dangerousness: A review and an empirical approach. British Journal of Social Work, 31, 249–269. Trocme, N., & Lindsey, D. (1996). What can child homicide rates tell us about the effectiveness of child welfare services? Child Abuse & Neglect, 20, 171–184. Vanamo, T., Kauppi, A., Karkola, K., Merikanto, J., & Rasanen, E. (2001). Intra-familial child homicide in Finland 1970–1994: Incidence, causes of death and demographic characteristics. Forensic Science International, 117, 199–204. 746 K. Cavanagh et al. / Child Abuse & Neglect 31 (2007) 731–746

Victoria Child Death Review Committee. (2004). Annual report. Melbourne, Australia. Wallace, A. (1986). Homicide: The social reality. NSW Bureau of Crime Statistics and Research, Attorney General’s Department, Research Study No. 5. Sydney. Wilczynski, A. (1995). Child killing by parents: A motivational model. Child Abuse Review, 4, 365–370. Wilczynski, A. (1997). Child homicide. London: Greenwich Medical Media. Wilson, R. F. (2006). Undeserved trust: Reﬂections on the ALI’s treatment of de facto parents. In R. F. Wilson (Ed.), Reconceiving the family: Critical reﬂections on the American Law Institute principles of the law on family dissolution (pp. 90–120). Cambridge: Cambridge University Press. Wheeler, P. (2003). Shaken baby syndrome—An introduction to the literature. Child Abuse Review, 12, 401–415.